def run_augment_ancestors(args):
setup_logging(args)
sample_data = tsinfer.SampleData.load(args.samples)
ancestors_trees = get_ancestors_trees_path(args.ancestors_trees,
args.samples)
output_path = args.augmented_ancestors
logger.info(
"Loading ancestral genealogies from {}".format(ancestors_trees))
ancestors_trees = tskit.load(ancestors_trees)
progress_monitor = ProgressMonitor(enabled=args.progress,
augment_ancestors=True)
# TODO Need some error checking on these values
n = args.num_samples
N = sample_data.num_samples
if n is None:
n = int(math.ceil(10 * N / 100))
sample_indexes = np.linspace(0, N - 1, num=n).astype(int)
ts = tsinfer.augment_ancestors(
sample_data,
ancestors_trees,
sample_indexes,
num_threads=args.num_threads,
path_compression=not args.no_path_compression,
progress_monitor=progress_monitor,
)
logger.info("Writing output tree sequence to {}".format(output_path))
ts.dump(output_path)
summarise_usage()
def run_augment(sample_data, ancestors_ts, subset, num_threads):
progress_monitor = tsinfer.cli.ProgressMonitor(enabled=True,
augment_ancestors=True)
return tsinfer.augment_ancestors(sample_data,
ancestors_ts,
subset,
num_threads=num_threads,
progress_monitor=progress_monitor)
def tsinfer_dev(n,
L,
seed,
num_threads=1,
recombination_rate=1e-8,
error_rate=0,
engine="C",
log_level="WARNING",
debug=True,
progress=False,
path_compression=True):
np.random.seed(seed)
random.seed(seed)
L_megabases = int(L * 10**6)
# daiquiri.setup(level=log_level)
ts = msprime.simulate(n,
Ne=10**4,
length=L_megabases,
recombination_rate=recombination_rate,
mutation_rate=1e-8,
random_seed=seed)
if debug:
print("num_sites = ", ts.num_sites)
assert ts.num_sites > 0
samples = tsinfer.SampleData.from_tree_sequence(ts)
ancestor_data = tsinfer.generate_ancestors(samples,
engine=engine,
num_threads=num_threads)
ancestors_ts = tsinfer.match_ancestors(samples,
ancestor_data,
engine=engine,
path_compression=True,
extended_checks=False)
ancestors_ts = tsinfer.augment_ancestors(samples,
ancestors_ts, [5, 6, 7],
engine=engine)
ts = tsinfer.match_samples(samples,
ancestors_ts,
path_compression=False,
engine=engine,
simplify=True)
# print(ts.tables.edges)
# print(ts.dump_tables())
# simplified = ts.simplify()
# print("edges before = ", simplified.num_edges)
# new_ancestors_ts = insert_srb_ancestors(ts)
# ts = tsinfer.match_samples(samples, new_ancestors_ts,
# path_compression=False, engine=engine,
# simplify=True)
# for tree in ts.trees():
# print(tree.interval)
# print(tree.draw(format="unicode"))
# print(ts.tables.edges)
# for tree in ts.trees():
# print(tree.draw(format="unicode"))
tsinfer.verify(samples, ts)